Plasma exosome-derived miRNA-887-5p alleviates high glucose- and lipid-induced endothelial cell dysfunction

Abstract

Background

Identifying differentially expressed miRNAs in plasma-derived exosomes associated with type 2 diabetes mellitus (T2DM) complicated by atherosclerosis (AS) and elucidating their roles in high-glucose/high-lipid-induced vascular endothelial dysfunction.

Methods

Plasma-derived exosomal miRNAs were isolated from people with T2DM complicated by AS and healthy controls. Followed by miRNA sequencing to characterize differential expression profiles between groups. GO and KEGG pathway enrichment analyses were performed on differentially expressed miRNAs. Human umbilical vein endothelial cells (HUVECs) were exposed to combined hyperglycemia and hyperlipidemia to establish an in vitro model of diabetic endothelial injury. HUVECs were subsequently transfected with miR-887-5p mimics, inhibitors or negative controls, and assessed for proliferation, migration, apoptosis, oxidative stress, and nitric oxide (NO) content.

Results

Sequencing revealed globally reduced plasma exosomal miRNA expression in people with T2DM and AS relative to healthy controls, with 21 upregulated and 23 downregulated miRNAs identified. Among these, hsa-miR-887-5p exhibited the greatest fold change of all detected miRNAs, while hsa-miR-96-5p and hsa-miR-183-5p (upregulated) and hsa-miR-410-3p (downregulated) harbored the most target genes implicated in diabetic atherosclerosis. Functionally, miR-887-5p enhanced HUVEC proliferation and migration under high-glucose/high-lipid conditions, elevated superoxide dismutase (SOD) activity, reduced lactate dehydrogenase (LDH) and malondialdehyde (MDA) levels, suppressed intracellular iNOS/NO and attenuated apoptosis.

Conclusion

Plasma exosomal miRNA expression is broadly reduced in people with T2DM complicated by AS. hsa-miR-887-5p, hsa-miR-96-5p, hsa-miR-183-5p, and hsa-miR-410-3p may emerge as candidates with diagnostic and therapeutic relevance in this context. Specifically, miR-887-5p mitigates high-glucose/high-lipid-induced vascular endothelial injury, warranting further investigation as a therapeutic target.

Introduction

Type 2 diabetes mellitus (T2DM) constitutes a growing global health burden, with the worldwide population of people with T2DM projected to exceed 693 million by 2045 [1]. Driven in part by shifting lifestyle patterns, its prevalence continues to rise, and its acute and chronic complications impose substantial societal costs [2]. Among these, Atherosclerosis (AS) is the most prevalent macrovascular complication of T2DM and the leading cause of mortality in this population [3, 4]. Sustained hyperglycemia and hyperlipidemia impair endothelial integrity, precipitating dysfunction that increases vascular permeability and facilitates transendothelial macrophage recruitment [5]. These macrophages subsequently accumulate lipids and transdifferentiate into foam cells, ultimately contributing to fibrous plaque formation [6–8].

Exosomes are nanoscale extracellular vesicles released upon cell activation or injury that serve as critical mediators of intercellular communication modulating the behavior of recipient cells [9]. They carry diverse cargo, including proteins, lipids, and miRNAs [10]. Of these, miRNAs, endogenous small non-coding RNAs, have attracted particular interest as post-transcriptional regulators of gene expression that can be shuttled via exosomes to both neighboring and distal cells [11, 12]. The seminal observation by Pegtel et al. that oncogenic viruses exploit exosomes to deliver functional miRNAs to healthy cells prompted the broader hypothesis that exogenous miRNAs or siRNAs could be packaged within exosomes for therapeutic gene delivery [13]. Despite considerable subsequent investigation across multiple disciplines, this field remains in its early stages [14]. Accumulating evidence links altered miRNA expression profiles to chronic vascular inflammation, including endothelial cell and macrophage dysfunction; people with diabetes exhibit disease-relevant shifts in miRNAs governing angiogenesis, vascular repair and endothelial homeostasis [15, 16]. Notably, miR-126 overexpression enhances the proangiogenic capacity of progenitor cells, the miR-146 family modulates oxidative stress and pro-inflammatory cytokine production, and reduced miR-146a expression has been implicated in the onset and progression of diabetic cardiovascular complications [17, 18]. Collectively, these observations underscore the therapeutic potential of miRNA-based strategies in diabetic vascular disease.

In the present study, next-generation sequencing (NGS) was employed to profile plasma exosomal miRNA expression in people with T2DM and AS, followed by bioinformatic analysis to identify key differentially expressed candidates. hsa-miR-887-5p was found to be among the most highly expressed exosomal miRNAs in this population. Using a HUVEC injury model of high-glucose/high-lipid exposure, we demonstrate that miR-887-5p rescues endothelial dysfunction, providing a potential rationale for its further evaluation as both a diagnostic biomarker and therapeutic target in T2DM with vascular complications.

Materials and methods

Reagents

Palmitic acid (PA) was obtained from Sigma Chemical (Cat. No. P0500, St. Louis, USA). VEGF was purchased from PeproTech (Cat. No. 450-32, New Jersey, USA). An inducible nitric oxide synthase (iNOS) enzyme immunoassay kit was procured from Jiangsu Meimian Industrial Co., Ltd. (Cat. No. MM-0889H2, Jiangsu, China), and DAF-FMDA (NO fluorescent probe) from Beyotime Biotechnology (Cat. No. S0019, Shanghai, China). Lactate dehydrogenase (LDH, Cat. No. A020-2), malondialdehyde (MDA, Cat. No. A003-1-2), and superoxide dismutase (SOD, Cat. No. A001-3-2) assay kits were purchased from Nanjing Jiancheng Biological Co. (Nanjing, China). A Cell Counting Kit-8 (CCK-8) was obtained from Biosharp (Cat. No. BS350A, Beijing, China), and an Annexin V-FITC/PI apoptosis kit from Sizhengbai Biotechnology Co., Ltd. (Cat. No. FXP018 Beijing, China).

General information

This study involved the collection and retrospective analysis of residual plasma samples from hospitalized patients undergoing routine clinical laboratory examinations. No extra blood collection was conducted, and written informed consent was obtained from all patients. The study was reviewed and approved by the Clinical Medical Research Ethics Committee of The First Affiliated Hospital of Anhui Medical University (Approval No. 20150192). Residual blood samples were collected from three people with T2DM complicated by AS, recruited from the Department of Endocrinology and three healthy controls from the Physical Examination Center of the First Affiliated Hospital of Anhui Medical University. Inclusion criteria were: (1) A diagnosis of T2DM (defined as progressive β-cell insulin secretory failure, frequently against a background of insulin resistance) for more than 5 years. (2) Poor glycemic control evidenced by classic hyperglycemia symptoms with a random plasma glucose ≥200 mg/dL (11.1 mmol/L) or fasting plasma glucose ≥126 mg/dL (7.0 mmol/L), where fasting was defined as no caloric intake for at least 8 h; (3) confirmed AS by systemic macrovascular ultrasound; and (4) age between 40 and 60 years. Exclusion criteria were: (1) other forms of diabetes or acute complications including malignant tumors, hypertension, active inflammation, polycystic ovary syndrome, or autoimmune diseases; (2) comorbidities affecting glucose metabolism, such as hyperthyroidism, acromegaly, or Cushing’s syndrome; and (3) severe physiological stress or serious cardiovascular or cerebrovascular disease.

Isolation and identification of exosomes

Participants fasted overnight for 8 h, venous blood was collected the following morning into EDTA-K2 anticoagulant tubes. Plasma was isolated by centrifugation at 3000 g at 4 °C for 10 min to remove any residual cell debris and platelets. Exosomes were subsequently extracted using the ExoQuick Exosome Precipitation Solution (System Biosciences, Mountain View, CA, USA). Briefly, 500 μL of plasma was mixed with 126 μL of ExoQuick Solution and incubated at 4 °C for 30 min. The mixture was then centrifuged at 1500 g for 30 min at room temperature, the supernatant discarded, and the pellet resuspended in 200 μL of phosphate-buffered saline (PBS). 20 μL of the resuspended exosome preparation was applied dropwise onto 200-mesh grids and incubated at room temperature for 10 min, negatively stained with 2% phosphotungstic acid for 3 min, and excess liquid was blotted with filter paper. Then, the exosomes were observed with a JEM1400 transmission electron microscope (TEM). Exosome size distribution was determined by nanoparticle tracking analysis (NTA).

Extraction of exosomal RNA

Total RNA was isolated using TRIzol reagent (Cat. No. 15596018CN, Life Technologies, Carlsbad, CA, USA) in conjunction with the miRNeasy Serum/Plasma Kit and an RNeasy MinElute column. Briefly, 200 μL of exosome suspension was combined with 1 mL TRIzol, and incubated at room temperature for 5 min, followed by the addition of 200 μL chloroform, vigorous shaking for 15 s, and a further 3-minute incubation. After centrifugation at 12,000×g (4 °C, 15 min), the aqueous phase was transferred to a clean tube, mixed with 900 μL absolute ethanol, and loaded onto an RNeasy MinElute column. Sequential washes were performed with 700 μL Buffer RWT, 500 μL Buffer RPE, and 500 μL 80% ethanol, each with centrifugation at ≥8000 × g for 15 s (2 min for the 80% ethanol wash). The membrane was dried for 5 min, and RNA was eluted in 15 μL of RNase-free water by full-speed centrifugation for 1 minute. RNA concentration was quantified using the Qubit RNA HS Assay.

miRNA library preparation

Small RNA libraries were constructed following standard protocols as previous described [19]. Total RNA was isolated using TRIzol reagent according to the manufacturer’s instructions. Briefly, small RNAs (18–28 nt) were size-selected from total RNA by 15% denaturing polyacrylamide gel electrophoresis, then subjected to sequential 5’ and 3’ adapter ligation using T4 RNA ligase, reverse transcription, and PCR amplification to generate cDNA libraries. Libraries were sequenced on an Illumina HiSeq 2500 platform in single-end 50 bp mode, yielding an average of 25 million raw reads per sample. All procedures and sequencing services were provided by Hefei Novell Gene Co., Ltd.

Next-generation sequencing of miRNAs

Sequencing was performed on the Illumina platform using reversible terminator-based synthesis, in which adapter-ligated cDNA fragments were immobilized on an optically transparent flow cell surface and amplified by bridge PCR to generate clusters of approximately 1000–6000 copies each, followed by four-color base incorporation for sequence readout. Raw 49-nt reads were processed using CPSS tools to remove adapter sequences, low-quality reads, and contaminants, yielding clean reads whose length distributions and inter-sample overlap were characterized. Clean reads were annotated to identify known miRNAs and determine their relative expression levels; sequences that could not be annotated against known databases were subjected to de novo miRNA prediction to identify candidate novel miRNAs. All sequencing data were deposited in the National Center for Biotechnology Information [NCBI] database (DATA number: PRJNA1139114).

Small RNA (sRNA) library data output statistics

Raw sequencing reads from the sRNA libraries were filtered according to the following quality control criteria: reads with low-quality score, those containing unknown bases (N content ≥10%), and those lacking a 3′ adapter sequence were discarded. The 3′ adapter was then trimmed from retained reads, and any sequences shorter than 15 nt or longer than 35 nt were excluded from downstream analysis.

Bioinformatics analysis

Differential miRNA expression profiles were clustered using Ward’s method to resolve miRNAs groupings and structural relationships. Target mRNAs were predicted using the R package multiMiR [20], which integrates both experimentally

validated targets from miRTarBase and computationally predicted targets from Targetscan, miRNA, and Microcosm databases. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were applied to the identified target genes to characterize their biological functions and associated signaling pathways. miRNAs-mRNAs interaction networks were constructed and visualized using Cytoscape (https://cytoscape.org/).

qRT-PCR validation

Plasma was collected from 10 healthy controls and 10 people with T2DM and AS per the inclusion criteria, and exosomes were isolated as described above. Exosomal miRNA was reverse-transcribed using the miRNA 1st Strand cDNA Synthesis Kit (Cat. No. MR101, Vazyme Biotech, Nanjing, China) per the manufacturer’s instructions. Quantitative RT-PCR was performed with the miRNA Universal SYBR qPCR Master Mix (Cat. No. MR102, Vazyme Biotech, Nanjing, China) using miRNA-specific forward primers and a universal reverse primer (Tsingke Biotechnology Co., Ltd., China). U6 small nuclear RNA served as the endogenous normalization control; primer sequences are provided in Table S4. All assays were performed across ten independent experiments.

Cell culture and high-glucose/high-lipid model

HUVECs were obtained from Procell Life Science & Technology Co., Ltd (Wuhan, China) and maintained in complete DMEM (10% fetal bovine serum, 1% 100 mg/L penicillin +100 mg/L streptomycin) at 37 °C and in a humidified 5% CO₂ atmosphere. To model diabetic endothelial injury in vitro, HUVECs were exposed to combined hyperglycemia and hyperlipidemia (25 mmol/L glucose and 200 μmol/L palmitic acid, termed HG + PA) [21]. VEGF (20 ng/mL) was included as a positive control given its established role in vascular endothelial cell regeneration [22, 23].

Cell transfection

HUVECs were transfected with miR-887-5p mimic, inhibitor, or negative control using the Lipofectamine™ 3000 Transfection Reagent according to the manufacturer’s instructions.

Cell proliferation, migration and LDH assay

Cells were seeded in 96-well plates, Using simple randomization, samples were assigned to experimental groups, ensuring unbiased and concealed grouping to reduce selection bias. They were allocated to six groups; normal control, HG + PA model, HG + PA+miR-887-5p mimic (100 nmol/L), HG + PA+miR-887-5p inhibitor (100 nmol/L), HG + PA+negative miR-887-5p (100 nmol/L), and HG + PA + VEGF (20 ng/mL). The investigator remained fully blinded to group allocation during both the experimental procedure and outcome assessment, effectively eliminating observer bias and assessment bias. Cell proliferation profiles of HUVECs were detected using a cell counting kit (CCK-8). After 24 h of culture, proliferation was assessed by adding 10 μL CCK-8 reagent per well protected from light, and measuring absorbance at 450 nm after 1 h. Migratory capacity was evaluated by scratch wound assay: monolayers were scratched with a sterile 200 μL pipette tip, cultured in complete medium for 24 h, and imaged using an inverted microscope; wound closure was quantified with ImageJ and expressed as percentage migration. LDH release was measured using the LDH assay kit per the manufacturer’s instructions. Six independent experiments were performed.

Tube formation assay

Matrigel™ was thawed overnight at 4 °C, and 100 μL was dispensed per well into pre-chilled 24-well plates on ice to ensure uniform coating and prevent premature gelation. Plates were then incubated at 37 °C for 1 h to allow solidification. HUVECs from each treatment group were seeded at 1.0 × 10⁵ cells per well and incubated at 37 °C with 5% CO₂. Capillary-like structure formation was assessed after 4 h by inverted light microscopy, and tube number was quantified with ImageJ. Three independent experiments were performed.

Intracellular NO detection

Intracellular NO levels were assessed as previously described [24]. Culture medium was changed to 4-amino-5-(methylamino)-2,7-difluorofluorescein diacetate (DAF-FMDA) fluorescence probe dilution, and the cells were incubated for 20 min before fluorescence microscopy imaging. Six independent experiments were conducted.

Enzyme-Linked Immunosorbent Assay (ELISA)

Intracellular iNOS content was measured using a commercially available ELISA kit per the manufacturer’s instructions [25]. Six independent experiments were conducted.

Malondialdehyde (MDA) and superoxide dismutase (SOD) assays

MDA and SOD levels in HUVECs were determined using the respective assay kits according to the manufacturer’s instructions, with absorbance read at 532 nm and 450 nm respectively. Six independent experiments were conducted.

Flow cytometry analysis

Apoptosis was assessed after 24 h of culture using an Annexin V-FITC/PI detection kit on a BD LSRFortessa flow cytometer (BD Biosciences, USA). Briefly, 100 µL of cell suspension was transferred to a 5 mL tube, incubated with 5 µL Annexin V-FITC for 5 min at room temperature in the dark, then supplemented with 10 µL PI and 200 µL 1× PBS before immediate acquisition and analysis using CELL Quest software. Six independent experiments were conducted.

Western blotting analysis

iNOS protein expression was assessed by Western blot. Total protein was extracted from HUVECs across treatment groups, resolved by SDS-PAGE, and transferred to a PVDF membrane. After blocking with 5% non-fat milk for 2 h at room temperature, membranes were probed overnight at 4 °C with anti-iNOS primary antibody (Cat. No. 22226-1-AP, Proteintech, Chicago, USA), washed three times with TBST, and incubated with HRP-conjugated secondary antibody (1:5000 in TBST) for 1 h at room temperature. Bands were visualized by ECL substrate and densitometrically quantified with ImageJ, with GAPDH serving as the loading control. Three independent experiments were performed.

Statistical analysis

All data are expressed as mean ± SD, with a minimum of three or six replicates. Statistical comparisons were made using a two-tailed Student’s t test or one-way ANOVA in GraphPad Prism. A P < 0.05 was considered statistically significant.

Results

Participant characteristics

Three people with T2DM complicated by AS and three healthy controls were enrolled. Clinical characteristics are summarized in Table S1.

Characterization of plasma-derived exosomes

Plasma-derived exosomes were characterized by TEM and NTA. TEM revealed homogeneous, spherical, membrane-bound vesicles consistent with canonical exosomal morphology (Fig. S1A), while NTA analysis demonstrated their size distribution and concentration, which displayed a unimodal profile with a peak centered at approximately 120 nm (Fig. S1B), collectively confirming successful exosome isolation.

Plasma exosomal miRNA profiles in T2DM with AS versus healthy controls

Exosomal miRNA reads were mapped to the hg19 reference genome. Global miRNA expression was reduced in plasma exosomes from people with T2DM and AS relative to healthy controls (Fig. 1A, B). Principal component analysis (PCA) showed that the first axis accounted for 45.3% of the variance, and the first two axes together explained 66.5% of the total inertia, with complete separation between the two groups (Fig. 1C).

Fig. 1. Total miRNA expression levels volcano plot, hierarchical clustering, and bioinformatic target gene analysis (n = 3).

A Box plots depicting the dispersion of miRNA expression across individual samples. B Histogram of total miRNA expression levels per sample. C PCA plot of plasma exosomal miRNAs from people with T2DM and AS versus healthy controls; color-coded groupings demonstrate complete separation between cohorts. D Volcano plot of all detected miRNAs, red indicates upregulated miRNAs (log₂(FC) > 1, P < 0.05), green indicates downregulated miRNAs (log₂(FC) < −1, P < 0.05), and black indicates no significant difference (P > 0.05). E Hierarchical clustering heatmap of exosomal miRNA expression patterns; red denotes increased expression (fold change >2) and blue denotes decreased expression (fold change <0.5). Venn diagrams of predicted target genes for upregulated (F) and downregulated (G) miRNAs across TargetScan, miRNA, Microcosm, and miRTarBase platforms; overlapping target genes were used for downstream GO, KEGG, and network analyses.

Identification of differentially expressed exosomal miRNAs and their target genes

Applying thresholds of |log₂(FC)| > 1 and P < 0.05, 44 differentially expressed miRNAs were identified in people with T2DM and AS versus healthy controls: 21 upregulated and 23 downregulated (Fig. 1D). Hierarchical clustering of expression profiles, where red denotes high and green denotes low expression, revealed tight intra-group clustering and clear inter-group separation (Fig. 1E). Target gene prediction was performed by integrating experimentally validated interactions from miRTarBase with computationally predicted targets from TargetScan, miRNA, and Microcosm, using the intersection of the three prediction databases as the predicted set. This yielded 4871 candidate target genes in total, 2454 for upregulated miRNAs (Fig. 1F) and 2417 for downregulated miRNAs (Fig. 1G), which were subsequently used for GO, KEGG, and network analyses. The top nine upregulated and downregulated miRNAs are listed in Tables S2 and S3, respectively. qRT-PCR validation of the top five upregulated and downregulated miRNAs, as well as those with the greatest number of target genes implicated in diabetic atherosclerosis, confirmed that miR-887-5p, miR-520g-3p, miR-6750-5p, miR-6738-3p, miR-6793-5p, and miR-96-5p were significantly elevated in patient plasma exosomes relative to healthy controls (P < 0.01), while miR-183-5p showed a non-significant trend toward upregulation (Fig. S1C). miR-494-5p, miR-4687-3p, miR-16-1-3p, miR-377-3p, miR-337-5p, and miR-410-3p were all significantly downregulated (P < 0.01; Fig. S1D). Among all candidates, miR-887-5p exhibited the greatest fold change, and qRT-PCR also detected low-level expression in healthy controls.

GO enrichment analysis of differentially expressed miRNA target genes

GO enrichment analysis of upregulated miRNA target genes revealed predominant involvement in protein catabolism and regulation of cell adhesion at the biological process level, with molecular functions centered on ubiquitin-like protein ligase binding and cadherin binding, both closely tied to endothelial integrity (Fig. 2A). Target genes of downregulated miRNAs were enriched in dephosphorylation, cell cycle regulation, and apoptotic signaling, processes directly implicated in high-glucose/high-lipid-induced endothelial injury (Fig. 2B). Collectively, these findings indicate that dysregulated exosomal miRNAs in T2DM with AS converge on pathways governing endothelial function, oxidative stress, and vascular inflammation.

Fig. 2. GO enrichment analysis and KEGG pathway and miRNA–mRNA network analyses of differentially expressed miRNA target genes.

A GO enrichment results for target genes of upregulated miRNAs, subdivided into biological process (a), molecular function (b), and cellular component (c). B GO enrichment results for target genes of downregulated miRNAs. In both panels, color intensity reflects −log₁₀(P-value) (red = more significant) and circle size represents the number of genes per GO term. C KEGG pathway analysis of exosomal miRNA target genes enriched in diabetes-related and vascular complication pathways. D miRNA–target gene interaction network constructed from differentially expressed miRNAs; target genes were derived from KEGG pathway analysis results.

KEGG pathway analysis and miRNA-mRNA network

KEGG analysis highlighted pathways directly relevant to diabetic macrovascular disease, including lipid metabolism and atherosclerosis, fluid shear stress and atherosclerosis, AGE-RAGE signaling in diabetic complications, and insulin signaling (Fig. 2C). Among the miRNAs mapping to these pathways, hsa-miR-96-5p, hsa-miR-183-5p (both upregulated) and hsa-miR-410-3p (downregulated) harbored a greater number of target genes associated with diabetic atherosclerosis, as visualized in the miRNA–mRNA interaction network constructed using Cytoscape (Fig. 2D). Notably, hsa-miR-887-5p exhibited the highest fold change of all detected miRNAs (Table S2), pointing to its potential diagnostic relevance and pathophysiological significance in diabetic vascular complications.

miR-887-5p promotes HUVEC proliferation, tube formation, and migration

To examine the functional role of miRNA-887-5p, we established a diabetic vascular endothelial injury model using HUVECs under HG + PA conditions. After 24 h, cells in the normal control group displayed the characteristic adherent, spindle-shaped, polygonal epithelioid morphology, whereas HG + PA-treated cells were reduced in number and shrunken in volume, confirming successful injury induction. miR-887-5p mimic and VEGF treatment restored cell number and maintained relatively regular morphology, while miR-887-5p inhibitor markedly reduced cell viability and disrupted cellular architecture; the negative miRNA-887-5p group was indistinguishable from the model group (Fig. 3A). CCK-8 assay confirmed that proliferative capacity was significantly impaired in the HG + PA model group relative to normal controls. miR-887-5p mimic and VEGF both rescued proliferation, while miR-887-5p inhibitor further suppressed it; the negative control had no significant effect (Fig. 3B). These results are consistent with cell morphology.

Fig. 3. miR-887-5p promotes HUVEC proliferation and migration.

A Representative brightfield micrographs of HUVEC morphology across treatment groups. Scale bar: 100 μm. B Statistics of CCK-8 assay. C Representative scratch wound images. Scale bar: 100 μm. D Quantification of wound closure (% migration). Data are presented as mean ± SD; n = 6. *P < 0.05, **P < 0.01 vs. control or model group; one-way ANOVA with Tukey’s post hoc test.

Tube formation assays corroborated these findings: angiogenic capacity was markedly enhanced by miR-887-5p mimic and VEGF relative to the model group, and further diminished by miR-887-5p inhibitor (Fig. S2), paralleling the proliferation results and indicating consistent regulatory effects of miR-887-5p on both processes. Scratch wound assays similarly showed that HG + PA inhibited HUVEC migration relative to controls, an effect significantly reversed by miR-887-5p mimic and VEGF but exacerbated by miR-887-5p inhibitor with no notable effect observed in the negative control group (Fig. 3C, D). Taken together, these data demonstrate that miR-887-5p exerts a broadly pro-angiogenic effect on HUVECs under diabetic injury conditions.

miR-887-5p attenuates high-glucose/high-lipid-induced oxidative stress in HUVECs

LDH release, a marker of cellular injury, was significantly elevated in the HG + PA model group relative to normal controls. miR-887-5p mimic and VEGF both reduced LDH content compared to the model group, while the negative control was comparable to the model group, and miR-887-5p inhibitor further increased LDH release (Fig. 4A), indicating that miR-887-5p mitigates HG + PA-induced cytotoxicity. Assessment of oxidative stress markers reinforced this conclusion: HG + PA caused a significant reduction in SOD activity and a marked elevation in MDA relative to normal controls. Compared with the model group, miR-887-5p mimic and VEGF restored SOD activity and lowered MDA, whereas miR-887-5p inhibitor had the opposite effect, and the negative control remained unchanged (Fig. 4B, C). These findings demonstrate that miR-887-5p protects against diabetic vascular endothelial injury, at least in part, by suppressing oxidative stress.

Fig. 4. miRNA-887-5p attenuates HG + PA-induced oxidative stress injury in HUVECs.

A LDH release. B SOD activity. C MDA content. Data are presented as the mean ± SD; n = 6. *P < 0.05, **P < 0.01 vs. control or model group; one-way ANOVA with Tukey’s post hoc test.

miR-887-5p reduces HG + PA-induced HUVEC apoptosis

Flow cytometric analysis revealed that HG + PA significantly increased the proportion of apoptotic HUVECs relative to normal controls. miR-887-5p mimic substantially reduced apoptosis compared with the model group, while miR-887-5p inhibitor further exacerbated it; the negative control showed no appreciable difference from the model group (Fig. 5). These results confirm that HG + PA-induced apoptosis in HUVECs is amenable to suppression by miRNA-887-5p.

Fig. 5. miRNA-887-5p reduces HG + PA-induced HUVEC apoptosis.

A Representative flow cytometry dot plots of Annexin V-FITC/PI staining. B Statistics of percentage apoptosis. Data are presented as the mean ± SD; n = 6. *P < 0.05, **P < 0.01 vs. control or model group; one-way ANOVA with Tukey’s post hoc test.

miR-887-5p alleviates HG + PA-induced HUVEC injury by downregulating iNOS/NO

Excessive NO production exerts cytotoxic and cytostatic effects and promotes cellular injury through oxidative stress mechanisms [26]. Consistent with this, NO levels were significantly elevated in the HG + PA model group compared with normal controls, an increase that was markedly attenuated by miR-887-5p mimic and further amplified by miR-887-5p inhibitor (Fig. 6A, B). iNOS, the principal enzymatic source of NO under inflammatory conditions, mirrored this pattern: iNOS content was significantly higher in the model group than in controls and was reduced by miR-887-5p mimic while being further elevated by miR-887-5p inhibitor (Fig. 6C). Western blot analysis corroborated these findings at the protein level, with iNOS protein markedly increased in the model group, significantly reduced by miR-887-5p mimic, and further elevated by miR-887-5p inhibitor (Fig. 6D). Collectively, these data indicate that miR-887-5p attenuates HG + PA-induced endothelial injury through suppression of the iNOS/NO axis.

Fig. 6. miR-887-5p reduces iNOS/NO levels in HG + PA-treated HUVECs.

A Representative fluorescence micrographs of DAF-FMDA-stained cells. Scale bar: 100 μm. B Quantification of mean fluorescence intensity. C Statistics of iNOS concentration; n = 6. D Representative western blot of iNOS (top) and corresponding statistical analysis (bottom) n = 3. All data are presented as mean ± SD. *P < 0.05, **P < 0.01 vs. control or model group; one-way ANOVA with Tukey’s post hoc test.

Discussion

AS remains the principal cause of disability and death among people with T2DM [27], making early intervention in diabetic vascular complications a central therapeutic priority.

Using NGS, we profiled plasma exosomal miRNAs in people with T2DM and AS versus healthy controls, identifying 44 differentially expressed candidates, 21 upregulated and 23 downregulated. Among these, hsa-miR-96-5p, hsa-miR-183-5p (upregulated) and hsa-miR-410-3p (downregulated) have harbored more of target genes relevant to diabetic AS. Overexpression of hsa-miR-96-5p has been shown to suppress glutathione, a key antioxidant, thereby promoting oxyradical and pro-inflammatory cytokine production [28]; conversely, intravenous glutathione antagonizes hyperglycemia-driven adhesion molecule upregulation and thrombin generation [29]. hsa-miR-183-5p is overexpressed in people with AS and drives vascular smooth muscle cell proliferation and migration [30]; separately, miR-183-5p enhances proliferation, invasion, and tube formation in HMEC-1 cells by suppressing FOXO1 [31]. miR-410-3p inhibits Toll-like receptor 2 signaling to alleviate mitochondrial dysfunction [32]. Mitochondrial dysfunction has been increasingly associated with the initiation and progression of AS by elevating the production of reactive oxygen species and mitochondrial oxidative stress damage [33].

Notably, among all detected miRNAs, hsa-miR-887-5p exhibited the greatest fold change. Sequencing detected it exclusively in the patient group, though qRT-PCR revealed trace-level expression in healthy controls. Its pronounced differential expression, apparent disease specificity, and novelty collectively identify it as the most likely upregulated exosomal candidate in diabetic atherosclerosis progression. This is expected to be a biological indicator of vascular complications in diabetic patients and it may provide a potential new target for the treatment for T2DM with AS. To interrogate its functional role, we transfected miR-887-5p mimics or inhibitor into HUVECs exposed to HG + PA. HG + PA increased LDH release, reduced cell viability and migration, and promoted apoptosis; miR-887-5p mimic reversed all of these effects, while the inhibitor exacerbated them, demonstrating that miR-887-5p acts as a cytoprotective factor against HG + PA-induced endothelial injury. High glucose and lipids are established inducers of endothelial apoptosis [34, 35]. The PI3K/Akt/eNOS axis is critical for endothelial survival and migration, and its impairment under diabetic conditions compromises vascular repair [36]. The ability of miR-887-5p to restore HUVEC viability and migration suggests a possible restoration of Akt signaling. Additionally, MAPK family members, namely p38, ERK1/2, and JNK, are stress-responsive kinases activated by hyperglycemia and hyperlipidemia to promote apoptosis and oxidative injury [37]; the reductions in LDH release and apoptosis observed with miR-887-5p may reflect inhibition of p38 or JNK activity. To our knowledge, this is the first demonstration that miR-887-5p ameliorates HG + PA-induced apoptosis in HUVECs, identifying a potential therapeutic avenue for diabetic microvascular complications.

Oxidative stress is a well-established driver of both micro- and macrovascular diabetic complications [38, 39]. Consistent with prior reports that people with T2DM exhibit elevated MDA and reduced SOD, patterns documented in the context of probiotic/synbiotic intervention studies and in people with diabetic foot [40, 41], we observed significantly decreased SOD activity and increased MDA in HG + PA-treated HUVECs, reflecting disrupted antioxidant defense. Notably, miR-887-5p mimic substantially elevated SOD activity and reduced MDA, attenuating oxidative stress and limiting HG + PA-induced cell damage. These effects of miR-887-5p on endothelial cells may contribute to the prevention of diabetic complications of vascular disease. The Nrf2/ARE pathway is a master regulator of endogenous antioxidant enzymes, including SOD [42], and the miR-887-5p-associated increase in SOD activity raises the possibility of Nrf2 involvement, warranting further investigation.

Under physiological conditions, endothelial NO is generated predominantly by eNOS and serves as a homeostatic regulator of vascular function. In T2DM, however, serum iNOS is markedly elevated, driving excessive NO production that triggers oxidative stress-mediated apoptosis and endothelial dysfunction [43–45].In line with these reports, HG + PA increased both NO and iNOS in HUVECs, and miR-887-5p mimic significantly attenuated both, implicating the iNOS/NO axis in the protective mechanism of miR-887-5p [46–48]. Whether this suppression is direct or mediated through intermediate targets remains to be resolved. The NF-κB pathway is a canonical driver of iNOS transcription, and hyperglycemia/hyperlipidemia-induced NF-κB activation in endothelial cells promotes iNOS overexpression, nitrosative stress, and apoptosis [49]; the reduction in iNOS protein by miR-887-5p mimic suggests possible NF-κB inhibition as an upstream mechanism. In the context of T2DM with AS, oxidative stress, inflammation, and apoptosis are deeply intertwined and mutually reinforcing, and miR-887-5p may confer systemic endothelial protection by targeting a shared upstream signaling node.

Several limitations should be acknowledged. The sequencing discovery cohort was small, and although candidate miRNAs were validated in an independent, larger cohort by qRT-PCR, studies with greater sample sizes are needed to consolidate these findings. Furthermore, functional investigation was limited to miR-887-5p in HUVECs; the other differentially expressed miRNAs identified here and the precise molecular mechanisms underlying miR-887-5p’s effects remain to be characterized.

Conclusion

We identified several differentially expressed plasma exosomal miRNAs in people with T2DM and AS, including hsa-miR-887-5p, hsa-miR-96-5p, hsa-miR-183-5p, and hsa-miR-410-3p, with potential diagnostic and therapeutic relevance. miRNA-887-5p in particular attenuated HG + PA-induced endothelial injury, positioning it as a candidate therapeutic target for people with T2DM and AS. Its disease-enriched expression pattern, highest fold change among all detected miRNAs, and cytoprotective function in endothelial cells support its further evaluation as a non-invasive early biomarker for vascular risk stratification in this population. Large-scale, multicenter prospective studies will nevertheless be required to establish its clinical sensitivity, specificity, and optimal diagnostic thresholds before translation to practice. The present findings provide a mechanistic and translational foundation for such efforts.

Author contributions

Y Chen, Q Zhang and B Shen designed the experiments and wrote the manuscript. Y Chen, F Xu, F Chen and H Xiao performed the experiments and analyzed the data. H Dong and J Chen contributed to the experiments. Y Zhou, Y Ma, and F Dai performed sample collection. F Xu, H Dong and F Chen conducted the related supplementary experiments. Y Chen, F Xu, Q Zhang, and B Shen supervised the project.

Funding

This study was supported by the National Natural Science Foundation of China under grant Nos. 82404621, 82574400, 32270590, and Clinical Research Development Special Fund of Anhui Hongde Shanyi Medical Development and Medical Assistance Foundation (KYZX2024012), the Science and Technology Development Fund, Macau SAR (File no. 0035/2024/RIA1, 0003/2025/NRP and 002/2023/ALC).

Data availability

The datasets supporting the conclusions of this article are included within the article and its additional files.

Competing interests

The authors declare no competing interests.

Ethical approval

All methods were performed in accordance with the relevant guidelines and regulations, and all experiments were approved by the Clinical Medical Research Ethics Committee of The First Affiliated Hospital of Anhui Medical University.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1038/s41387-026-00451-9.